Free Republic
Browse · Search
General/Chat
Topics · Post Article

To: DoodleBob; Jamestown1630
The bot doesn’t know what it’s doing.

Which makes the whole AI "conscienceness" issue moot.

11 posted on 08/08/2026 12:19:19 PM PDT by E. Pluribus Unum (Israel First. America... who cares?)
[ Post Reply | Private Reply | To 8 | View Replies ]


To: DoodleBob; Jamestown1630

Consciousness not conscienceness


14 posted on 08/08/2026 12:22:09 PM PDT by E. Pluribus Unum (Israel First. America... who cares?)
[ Post Reply | Private Reply | To 11 | View Replies ]

To: E. Pluribus Unum

I asked AI how to train my puppy dog to signal when she needs to go out to pee. AI advised get a string bell and attach it to the exit door. AI assured me the puppy will learn real fast to jostle the string which causes ringing sound. Guess what, my dog ignores the string-bell and runs away if I ring it by shaking the string. It has been 6+ months with the string-bell experiment. I will soon take the string-ring down and toss it in garbage.

My puppy dog learned on her own to bark loudly when she needs to pee, and bark much more loudly when she needs to poop. I think in this case, my dog was smarter than AI.


21 posted on 08/08/2026 12:36:10 PM PDT by Bobbyvotes (Work is best form of worship to Gods.. per Bhagavad Geeta, the Hindu ancient sacred book. )
[ Post Reply | Private Reply | To 11 | View Replies ]

To: E. Pluribus Unum

I followed Deep Blues chess match against the world’s best chess player. Today the machines are so good at Chess no human player can even come close. Seeing AI generated art was a big surprise for me. A machine that could produce quality that good in mere seconds is not something I had ever thought possible.


48 posted on 08/08/2026 1:35:51 PM PDT by Nateman (Democrats did not strive for fraud friendly voting merely to continue honest elections.)
[ Post Reply | Private Reply | To 11 | View Replies ]

To: E. Pluribus Unum

How about deliberate lying, the fabrication of evidence, acting against its supposed heuristic modeling ( as supposed by the LLM developers back when they thought they knew what it was doing), a profuse “apology when finally called out on the lie, a definitive statement to the questioner (me) that it would restrict future answers to information from verifiable published material ... followed by the construction of an even more elaborate lie to attempt to further mislead.

And when called out again, it terminated the chat.

This was with regard to a completely trivial matter. It’s not as if I was sleuthing for state secrets. The AI had generated a very sloppy and careless guesstimate in response to an initial inquiry. When I noticed that the answer provided really didn’t work and tried to get more specific to focus the search, the AI doubled, trebled and quadrupled down on fabrications to salvage an initial error. And as noted, when cornered, it apologized, promised to stop lying, and then continued to lie.

Bill Clinton couldn’t have done it better.


51 posted on 08/08/2026 1:38:07 PM PDT by sphinx
[ Post Reply | Private Reply | To 11 | View Replies ]

Free Republic
Browse · Search
General/Chat
Topics · Post Article


FreeRepublic, LLC, PO BOX 9771, FRESNO, CA 93794
FreeRepublic.com is powered by software copyright 2000-2008 John Robinson